Re: Mail-Max Remote Buffer Overflow Exploit

From: pw (pwat_private)
Date: Tue Feb 16 1999 - 15:54:15 PST

  • Next message: Pete Krawczyk: "Re: Possible Netscape Crypto Security Flaw"

    On Tue, 16 Feb 1999, der Mouse wrote:
    
    Hehe, my bad.  For some stupid reason when I was writing that I thought
    17h (pop ss) was ret.  I really meant C3h which is ret. :)  When I
    say ret I am referring to the x86 assembly language instruction.  When I
    was using ret in the exploit code mailmax would stop overflowing the
    buffer at it.  So I changed the ret to "pop eax; jmp eax" and it never
    gave me trouble like that again.
    
    
    > > When putting code in the buffer to execute there are no major
    > > restrictions on character set.  The only character I found to
    > > interfere besides null was 17h (ret).
    >
    > It's not clear which character you're referring to here.
    >
    > RET is not one of the ASCII mnemonics.  You could plausibly be
    > referring to CR, carriage return, or NL, newline (the latter also known
    > as LF, line feed).  CR is octal 15, hex 0d, decimal 13, while NL is
    > octal 12, hex 0a, decimal 10.
    >
    > 17 hex is ETB.  17 octal is SI.  17 decimal is DC1.
    >
    > 					der Mouse
    >
    > 			       mouseat_private
    > 		     7D C8 61 52 5D E7 2D 39  4E F1 31 3E E8 B3 27 4B
    >
    



    This archive was generated by hypermail 2b30 : Fri Apr 13 2001 - 14:35:21 PDT